Computing Cohabitation: A Framework For Indexing More-Than-Human Streetscapes

Contemporary streetscapes contribute to the decline of avian species that have adapted over time to rely on the built environment for nesting and reproduction. Conservation measures include the placement of artificial nests on architectural facades; however, the future success of these nesting sites depends on a combination of natural and built features that mediate interspecies dynamics. This paper proposes a computational framework for predicting viable nesting sites for the western house martin (Delichon urbicum), currently classified as Near Threatened (NT) in parts of Europe by the International Union for Conservation (IUCN). By leveraging the use of Street View Imagery (SVI) and deep learning algorithms, the proposed method assesses more-than-human habitat along streetscapes and automates the identification of potential nesting sites. The paper discusses the importance of developing tools for aligning design practice with wildlife conservation and presents opportunities for integrating a more-than-human perspective to urban renovation agendas.